Original code for bioinformatics analysis of ATAC-seq & RNA-seq data

A shared NF-κB-driven senescence response shows distinct temporal dynamics under CDK4/6 inhibition and doxorubicin treatment in liposarcoma

Authors: Joanna Lan-Hing Yeung,1,3 Justin Rendleman,1,3 Nicole Pagane,1 Lauren Anderson,1 Irene Duba,1 Ria Hosuru,1 Bat-Ider Tumenbayar,1 Andrew Koff,2 Viviana I. Risca1,4,*

1Laboratory of Genome Architecture and Dynamics, The Rockefeller University, New York, NY 10065, USA. 2Molecular Biology Program, Memorial Sloan Kettering Cancer Center, New York, NY 10065, USA. 3These authors contributed equally 4Lead contact *Correspondence: vrisca@rockefeller.edu

SUMMARY

Cancer therapies can trigger geroconversion in tumor cells which influences patient outcomes, however it remains unclear how the characteristics of therapy-induced senescent states vary between treatments. While senescence is generally initiated by some form of DNA damage, CDK4/6 inhibitors (CDK4/6i) are unique in that they cause geroconversion in a damage-independent manner. Here, we use a multi-omics approach to profile the temporal dynamics of transcriptional rewiring during therapy-induced senescence in CDK4/6i-treated liposarcoma cells and compare this to the chemotherapy drug doxorubicin, which generates double-stranded breaks prompting geroconversion. We found the dynamics of senescence-related phenotypes varies between therapies, largely explained by a DNA damage-dependent signature specific to doxorubicin. Inevitably, there emerged a shared senescence-associated secretory phenotype (SASP) marked by NF-κB activation, coinciding with stable arrest. Pharmacological inhibition of NF-κB suppressed upregulation of this shared SASP program, confirming its regulatory role even in the absence of DNA damage.

Directory tree

├── Figure2
│   └── Figure2.Rmd 
├── Figure3
│   ├── deepTools_files
│   │   └── deeptools_Coverage_1kb.slurm
│   └── Figure3.Rmd
├── Figure4
│   ├── Figure4.Rmd
│   └── motif_analysis_input
│       └── MEME_AME.slurm
├── Figure5
│   └── Figure5.Rmd
├── Figure6
│   ├── deepTools_files
│   │   └── deeptools_Coverage_1kb.slurm
│   ├── Figure6.Rmd
│   └── NFKB_tp_peaks
│       └── get_tp_peaks_motifs.slurm
├── FigureS1
│   └── FigureS1.Rmd
├── FigureS2
│   ├── deepTools_files
│   │   └── deeptools_Coverage_1kb.slurm
│   ├── E2F_tp_peaks
│   │   └── get_tp_peaks_motifs.slurm
│   └── FigureS2.Rmd
├── FigureS3
│   ├── FigureS3.Rmd
│   └── motif_analysis_input
│       └── MEME_AME.slurm
├── FigureS4
│   └── FigureS4.Rmd
├── FigureS5
│   ├── FigureS5.Rmd
│   └── Supplemental_Figure_SASP_composition.Rmd
├── FigureS6
│   ├── FigureS6.Rmd
│   └── motif_analysis_input
│       └── MEME_AME.slurm
├── generate_script_tree.sh
├── Preprocessing_scripts
│   ├── Assess_QC_metrics
│   │   ├── calculate_FRIP_R_script.R
│   │   ├── calculate_FRIP_scores_R.slurm
│   │   ├── cat_duplication_metrics.sh
│   │   ├── cat_PBC_QC.sh
│   │   ├── count_mito_readpairs.sh
│   │   ├── get_total_reads.sh
│   │   ├── run_array_job.sh
│   │   └── tss_enrichment_score.sh
│   ├── ATACseq
│   │   ├── atac_1_fastq_trim.slurm
│   │   ├── atac_2_fastq_align.slurm
│   │   ├── atac_3_bam_managemito.slurm
│   │   ├── atac_4_bam_markduplicates.slurm
│   │   ├── atac_5_bam_FLD.slurm
│   │   ├── atac_6.3_IDR_tagalign.slurm
│   │   ├── atac_6_IDR_tagalign_2reps.slurm
│   │   ├── bamCoverage_scaleFactor_noOffset.slurm
│   │   ├── cutsCoverage.R
│   │   ├── cutsCoverage.slurm
│   │   ├── iterative_peak_filtering
│   │   │   ├── mergedFW_peakCall.R
│   │   │   └── peakfiltering.slurm
│   │   └── mergebam.slurm
│   └── RNAseq
│       └── kallisto_counting.slurm
└── Supplemental_Tables
    └── Supplemental_Tables.Rmd

Description of scripts

• ./Figure2/Figure2.Rmd
  ⤷ code to generate figures for Figure 2

• ./Figure3/deepTools_files/deeptools_Coverage_1kb.slurm
  ⤷ code to plot Tn5 insertion sites around p53 motifs

• ./Figure3/Figure3.Rmd
  ⤷ code to generate figures for Figure 3

• ./Figure4/Figure4.Rmd
  ⤷ code to generate figures for Figure 4

• ./Figure4/motif_analysis_input/MEME_AME.slurm
  ⤷ *code to perform motif enrichment analysis using MEME-AME HOCOMOCO v11 database *

• ./Figure5/Figure5.Rmd
  ⤷ code to generate figures for Figure 5

• ./Figure6/deepTools_files/deeptools_Coverage_1kb.slurm
  ⤷ code to plot Tn5 insertion sites around NF-kB motifs

• ./Figure6/Figure6.Rmd
  ⤷ code to generate figures for Figure 6.

• ./Figure6/NFKB_tp_peaks/get_tp_peaks_motifs.slurm
  ⤷ code to count the number of true positive peaks for NFKB family of motifs after MEME-AME analysis.

• ./FigureS1/FigureS1.Rmd
  ⤷ code to generate figures for Figure S1.

• ./FigureS2/deepTools_files/deeptools_Coverage_1kb.slurm
  ⤷ code to plot Tn5 insertion sites around E2F motifs

• ./FigureS2/E2F_tp_peaks/get_tp_peaks_motifs.slurm
  ⤷ code to count the number of true positive peaks for E2F family of motifs after MEME-AME analysis.

• ./FigureS2/FigureS2.Rmd
  ⤷ code to generate figures for Figure S2

• ./FigureS3/FigureS3.Rmd
  ⤷ code to generate figures for Figure S3

• ./FigureS3/motif_analysis_input/MEME_AME.slurm
  ⤷ code to perform motif enrichment analysis using MEME-AME HOCOMOCO v11 database

• ./FigureS4/FigureS4.Rmd
  ⤷ code to generate figures for Figure S4

• ./FigureS5/FigureS5.Rmd
  ⤷ code to generate figures for Figure S5

• ./FigureS6/FigureS6.Rmd
  ⤷ code to generate figures for Figure S6

• ./FigureS6/motif_analysis_input/MEME_AME.slurm
  ⤷ code to perform motif enrichment analysis using MEME-AME HOCOMOCO v11 database

• ./Preprocessing_scripts/Assess_QC_metrics/calculate_FRIP_R_script.R
  ⤷ calculate FRIP score for each individual sample's narrowPeak & bam file in R

• ./Preprocessing_scripts/Assess_QC_metrics/calculate_FRIP_scores_R.slurm
  ⤷ slurm script to run ./Preprocessing_scripts/Assess_QC_metrics/calculate_FRIP_R_script.R

• ./Preprocessing_scripts/Assess_QC_metrics/cat_duplication_metrics.sh
  ⤷ *script to extract duplication metrics based on the output metrics.txt files generated from /Preprocessing_scripts/ATACseq/atac_4_bam_markduplicates.slurm

• ./Preprocessing_scripts/Assess_QC_metrics/cat_PBC_QC.sh
  ⤷ *script to extract duplication metrics based on the output PBC_QC.txt files generated from /Preprocessing_scripts/ATACseq/atac_4_bam_markduplicates.slurm

• ./Preprocessing_scripts/Assess_QC_metrics/count_mito_readpairs.sh
  ⤷ *script to count the number of mitochondrial read pairs based on the output *.flagstat & non.Mito.flagstat files generated from ./Preprocessing_scripts/ATACseq/atac_3_bam_managemito.slurm

• ./Preprocessing_scripts/Assess_QC_metrics/get_total_reads.sh
  ⤷ *script to extract total raw read pairs based on the output .log files generated from ./Preprocessing_scripts/ATACseq/atac_2_fastq_align.slurm

• ./Preprocessing_scripts/Assess_QC_metrics/run_array_job.sh
  ⤷ script to run ./Preprocessing_scripts/Assess_QC_metrics/tss_enrichment_score.sh

• ./Preprocessing_scripts/Assess_QC_metrics/tss_enrichment_score.sh
  ⤷ script to calculate TSS enrichment scores based on ENCODE ATAC-seq standards and using Gencode v41 gene annotation file

• ./Preprocessing_scripts/ATACseq/atac_1_fastq_trim.slurm
  ⤷ script to trim off Tn5 adapters from fastq files

• ./Preprocessing_scripts/ATACseq/atac_2_fastq_align.slurm
  ⤷ script to align reads from trimmed fastqs to hg38 genome & generate bam files

• ./Preprocessing_scripts/ATACseq/atac_3_bam_managemito.slurm
  ⤷ script to remove mitochondrial reads

• ./Preprocessing_scripts/ATACseq/atac_4_bam_markduplicates.slurm
  ⤷ script to generate final bam files and remove duplicate reads

• ./Preprocessing_scripts/ATACseq/atac_5_bam_FLD.slurm
  ⤷ *Script that runs picard CollectInsertSizeMetrics to generate the fragment length distribution *

• ./Preprocessing_scripts/ATACseq/atac_6.3_IDR_tagalign.slurm
  ⤷ script to call peaks with MACS2 and perform IDR on samples with 3 technical replicates.

• ./Preprocessing_scripts/ATACseq/atac_6_IDR_tagalign_2reps.slurm
  ⤷ script to call peaks with MACS2 and perform IDR on samples with 2 technical replicates.

• ./Preprocessing_scripts/ATACseq/bamCoverage_scaleFactor_noOffset.slurm
  ⤷ script to generate bigwigs normalized by sequencing depth from bam files

• ./Preprocessing_scripts/ATACseq/cutsCoverage.R
  ⤷ script that extracts the ends of reads after adjusting for Tn5 insertion bias and converting it to a normalized bigwig of Tn5 insertion sites to be used as input for deeptools_Coverage_1kb.slurm

• ./Preprocessing_scripts/ATACseq/cutsCoverage.slurm
  ⤷ script that runs ./Preprocessing_scripts/ATACseq/cutsCoverage.R

• ./Preprocessing_scripts/ATACseq/iterative_peak_filtering/mergedFW_peakCall.R
  ⤷ after performing IDR and concatenating all narrowPeak files into a master redundant peak set, this script iteratively filters overlapping peaks

• ./Preprocessing_scripts/ATACseq/iterative_peak_filtering/peakfiltering.slurm
  ⤷ script that runs ./Preprocessing_scripts/ATACseq/iterative_peak_filtering/mergedFW_peakCall.R

• ./Preprocessing_scripts/ATACseq/mergebam.slurm
  ⤷ script to merge bam files from the same condition

• ./Preprocessing_scripts/RNAseq/kallisto_counting.slurm
  ⤷ script to count transcript abundance using kallisto

• ./Supplemental_Tables/Supplemental_Tables.Rmd
  ⤷ code to generate supplemental tables